The difference between oblimin and promax rotation in EFA

[fouad alhaydari]

Hi dear,
I have a problem in recognizing of using oblimin or promax rotation in factor analysis with IBM statistics SPSS 20, and my samples are 50 person.
thank you for replying

[ben earnhart]

Your question is rather vague, and refers to SPSS, not Stata. But I'll interpret your question to be "should I use oblimin or promax?"

Given that you have already decided to use an oblique solution, and that it's an EFA, well, try them both. They should end up with similar factor structures. If they don't:

• You might have variables that really don't belong together. Apart from the obvious, consider measurement scales and directionality..
• Does one solution make more substantive sense than the other?
• Can you get another sample and do a CFA instead of EFA?

[fouad alhaydari]

Dear Ben

Thanks for your prompt response.

I`d like to confirm the following:
- The variables used in my study are correlated
- The number of observations is small (50)

I `d be appreciated if you could advice whether oblimin or promax is appropriate to analyse the data.

kind regards

Fouad Alhaydari

[ben earnhart]

Have you tried them both, and do they give you different results? If your structure is stable, I'd expect them to give similar results. How many variables do you have and how many factors are you extracting?

[fouad alhaydari]

Thanks Ben
Thanks again for your response.
I confirm that I have used both the promax and oblimin and got different results for otherv ariables. i have used 20 variables, and extracted 6 factors.
kind regards
Fouad Alhaydari